This grant is to solve ab initio the phase problem, a classic high risk/high impact challenge in structural biology. The macromolecular structure determination will be improved by novel approaches to identifying non-crystallographic symmetry, to direct methods an to phase extension. The first target will be the most frequent, and often the most challenging, cases in macromolecular crystallography: those of the molecules in the crystal lattice having and additional, non-crystallographic symmetry. The proposal is to develop the initial implementation of: - Identification of the rotational component of non-crystallographic symmetry by improving the self-rotation function; - A novel method to identify the molecule's shape and position ab initio; - Efficient use of the a priori information about the electron density inside the mask; - Ab initio phasing up to the intermediate resolution by conjugated gradient method; - Phase extension by state-of-the art solvent-flattening and NCS-averaging methods. These methods will be validated with a representative range of samples. It is expected that the developed framework will form a basis for the generally valid subsequent application that will led the structural biologists skip laborious and uncertain stage of developing heavy atom derivatized crystals. In terms of understanding the molecular biology, the highest expected impact is in structure determination of complex molecular assemblies. [unreadable] [unreadable]